Biology subjects

Li, A. J.

Publications and source records attributed to Li, A. J..

10 recordsLinked to original sources

bioRxiv2025

All-atom protein design via SE(3) flow matching with ProteinZen

The advent of generative models of protein structure has greatly accelerated our ability to perform de novo protein design, especially concerning design at coarser physical scales such as backbone generation and protein binder design. How-ever, the design of precise placements at atomic scales remains a challenge for existing design methods. One avenue towards higher fidelity atomic-scale design is via generative models with full atomic resolution, but is complicated by the intricacies of simultaneously designing both discrete protein sequence and continuous atomic positionings. In this work we propose a framework to capture this interplay by decomposing residues into collections of oriented rigid bodies, allowing us to apply SE(3) flow-matching for all-atom protein structure generation. Our method, ProteinZen1, generates designs with high sequence-structure consistency while retaining competitive diversity and novelty on both unconditional and conditional generation tasks. We demonstrate competitive performance for unconditional monomer design and state-of-the-art performance on various forms of motif scaffolding, including full-atom motif scaffolding and motif scaffolding without specifying motif segment spacing or relative sequence order.

bioengineering↗

bioRxiv2025

The mutational landscape defines the proteome and spatial organization of tumor, stroma and immune cells in ovarian cancer

High-grade serous ovarian cancer (HGSOC) is a highly aggressive and lethal form of ovarian cancer. Challenges to diagnosis and treatment include a lack of effective screening methods for early detection, the absence of cancer-specific symptoms, and the development of chemoresistance. The genomic instability of HGSOC, further complicated by homologous recombination deficiency (HRD), leads to heterogeneity in HGSOC tumors and patient response to treatment. This makes it challenging to develop a single, effective diagnostic and treatment approach for this disease. Proteogenomic studies have provided some insight into HGSOC biology, but a deeper understanding of the tumor proteome through chemotherapy and disease recurrence is needed. Here we have profiled the proteome of tumors from 29 HGSOC patients before and after multiple rounds of chemotherapy. The proteome of HGSOC tumors remained unchanged for individual patients even after numerous rounds of chemotherapy. Differential expression analysis revealed known and novel proteins associated with chemoresistance and HRD status, further supported by similar changes at the genetic and epigenetic levels. We found that HRD affected proteins related to immune pathways. HRD was also associated with more shared T cell receptor (TCR) CDR3 repertoires of tumor-infiltrating T cells and increased neoantigen counts. ERAP1, a protein involved in peptide trimming before antigen presentation to immune cells via MHC-class I, was overexpressed in homologous recombination proficient (HRP) tumors and negatively correlated to neoantigen count. Its potential role in immune suppression in HRP tumors makes it an attractive therapeutic target that may be effective in combination with immunotherapy, particularly for HRP tumors. 26-plex immunostaining of HGSOC whole tissues further revealed significant differences in the spatial proximities of immune cells to each other and to tumor cells based on HRD status. Through proteomic and imaging analysis, this work has shown that the immune landscape of HGSOC tumors is influenced by homologous recombination status and identified candidate drivers of HGSOC biology.

cancer biology↗

bioRxiv2025

The mutational and clonality profile of HGSOC is established early in tumor development and conserved throughout therapy resistance

High grade serous ovarian cancer (HGSOC) is the most lethal gynecologic malignancy, killing more than 9,000 women each year in the United States alone. Nearly 80% of patients with HGSOC tumors will experience recurrence within 5 years, but little is known about the mechanisms that drive this process. Intratumor heterogeneity is believed to be a key feature of recurrence and resistance in HGSOC tumors, with early studies reporting diverse and complex mechanisms of the seeding of metastatic sites, including metastasis reseeding the primary tumor. Few studies have investigated temporal changes to clonality and structural variants through disease recurrence and the development of chemoresistance as surgical debulking is not frequently performed at this later stage. We performed multi-omic profiling in paired chemo-naive and chemoresistant tumors from 32 HGSOC patients to investigate the mutational and genomic landscape of disease progression. Somatic mutation profiles were largely conserved through disease progression. Mutational burdens did not significantly differ across recurrence but were driven by homologous recombination repair deficiency status. Clonal composition and dynamics were measured through variant allele frequency alterations as tumors progressed from primary chemo-naive to recurrent chemo-resistant tumors. A novel candidate driver gene, MDC1, from the homologous recombination repair pathway, was significantly mutated and over-represented in patients with homologous recombination proficient tumors, with somatic mutations clustered in a single exon. Tumor evolution and phylogeny revealed that few changes in clonal abundance and complexity occur across the disease course in these patients. Taking structural variants into account, homologous recombination repair proficient (HRP) tumors tend to be polyclonal while homologous recombination repair deficient (HRD) tumors tend to be monoclonal, accompanied by a longer progression-free survival than the HRP patients. Three distinct classes of tumors were identified by structural variant signature analysis: tumors defined by DNA losses, tumors defined by DNA gains, and tumors defined by copy number neutral changes, which were largely defined by HRD status. Each class displayed distinct regions of the chromosome that were frequently affected by large scale SV events (>5Mb). Although no regions were frequently altered in recurrent tumors, GO analysis revealed that recurrent tumors have a significantly reduced immune response, which was not seen in the primary tumors. Ultra long read sequencing validated a majority of the SVs identified in short read sequencing and identified additional SVs undetected by short reads. These analyses identify that the phenotype of high grade serous ovarian tumors as defined by mutation and clonality profiles is established early in disease development and remain largely unchanged through chemotherapy and recurrence. This, when considered with the significant inter-patient heterogeneity identified in HGSOC, demonstrates the need for personalized therapies based on tumor profiling.

genomics↗

bioRxiv2024

Simultaneous mesoscopic measurement and manipulation of mouse cortical activity

Dynamics of activity across the cerebral cortex at the mesoscopic scale -- coordinated fluctuations of local populations of neurons -- are essential to perception and cognition and relevant to computations like sensorimotor integration and goal-directed task engagement. However, understanding direct causal links between population dynamics and behavior requires the ability to manipulate mesoscale activity and observe the effect of manipulation across multiple brain regions simultaneously. Here, we develop a novel system enabling simultaneous recording and manipulation of activity across the dorsal cortex of awake mice, compatible with large-scale electrophysiology from any region across the brain. Transgenic mice expressing the GCaMP calcium sensor are injected systemically with an adeno-associated virus driving expression of the ChrimsonR excitatory opsin. This strategy drives expression of the blue-excited calcium indicator, GCaMP, in excitatory neurons and red-excited Chrimson opsin in inhibitory neurons. We demonstrate widefield single-photon calcium imaging and simultaneous galvo-targeted laser stimulation over the entire dorsal cortical surface. The light channels of the imaging and the opsin do not interfere. We characterize the spatial and temporal resolution of the method, which is suitable for targeting specific cortical regions and specific time windows in behavioral tasks. The preparation is stable over many months and thus well-suited for long-term behavioral experiments. This technique allows for studying the effect of cortical perturbations on cortex-wide activity, on subcortical spiking activity, and on behavior, and for designing systems to control cortical activity in closed-loop.

neuroscience↗

bioRxiv2023

CB1R blockade unmasks TRPV1-mediated contextual fear generalization in female, but not male rats

Increasing evidence suggests that the neurobiological processes that govern learning and memory can be different in males and females, and here we asked specifically whether the endocannabinoid (eCB) system could modulate Pavlovian fear conditioning in a sex-dependent manner. Systemic (i.p.) injection of CB1R antagonist AM251 in adult male and female Sprague Dawley rats prior to auditory cued fear conditioning produced a female-specific increase in freezing that persisted across extinction and extinction retrieval tests but was prevented by co-administration of TRPV1R antagonist Capsazepine. Notably, AM251 also produced robust freezing in a novel context prior to auditory cue presentation the day following drug administration, but not the day of, suggesting that CB1R blockade elicited contextual fear generalization in females. To identify a potential synaptic mechanism for these sex differences, we next used liquid chromatography/tandem mass spectrometry, Western Blot, and confocal-assisted immunofluorescence techniques to quantify anandamide (AEA), TRPV1R, and perisomatic CB1R expression, respectively, focusing on the ventral hippocampus (vHip). Fear conditioning elicited increased vHip AEA levels in females only, and in both sexes, CB1R expression around vHip efferents targeting the basolateral amygdala (BLA) was twice that at neighboring vHip neurons. Finally, quantification of the vHip-BLA projections themselves revealed that females have over twice the number of neurons in this pathway that males do. Together, our data support a model in which sexual dimorphism in vHip-BLA circuitry promotes a female-specific dependence on CB1Rs for context processing that is sensitive to TRPV1-mediated disruption when CB1Rs are blocked.

neuroscience↗

bioRxiv2022

Excitation creates a distributed pattern of cortical suppression due to varied recurrent input

Dense local, recurrent connections are a major feature of cortical circuits, yet how they affect neurons responses is unclear, with some studies reporting weak recurrent effects, some amplification, and others showing instead local suppression. Here, we show that optogenetic input to mouse V1 excitatory neurons generates salt-and-pepper patterns of both excitation and suppression. Responses in individual neurons are not strongly predicted by that neurons direct input. A balanced-state network model reconciles a set of diverse observations: the observed dynamics, suppressed responses, decoupling of input and output, and long tail of excited responses. The model shows recurrent excitatory-excitatory connections are strong and also variable across neurons. Together, these results demonstrate that excitatory recurrent connections can have major effects on cortical computations, by shaping and changing neurons responses to input.

neuroscience↗

bioRxiv2022

Neural Network-Derived Potts Models for Structure-Based Protein Design using Backbone Atomic Coordinates and Tertiary Motifs

AO_SCPLOWBSTRACTC_SCPLOWDesigning novel proteins to perform desired functions, such as binding or catalysis, is a major goal in synthetic biology. A variety of computational approaches can aid in this task. An energy-based framework rooted in the sequence-structure statistics of tertiary motifs (TERMs) can be used for sequence design on pre-defined backbones. Neural network models that use backbone coordinate-derived features provide another way to design new proteins. In this work, we combine the two methods to make neural structure-based models more suitable for protein design. Specifically, we supplement backbone-coordinate features with TERM-derived data, as inputs, and we generate energy functions as outputs. We present two architectures that generate Potts models over the sequence space: TERMinator, which uses both TERM-based and coordinate-based information, and COORDinator, which uses only coordinate-based information. Using these two models, we demonstrate that TERMs can be utilized to improve native sequence recovery performance of neural models. Furthermore, we demonstrate that sequences designed by TERMinator are predicted to fold to their target structures by AlphaFold. Finally, we show that both TERMinator and COORDinator learn notions of energetics, and these methods can be fine-tuned on experimental data to improve predictions. Our results suggest that using TERM-based and coordinate-based features together may be beneficial for protein design and that structure-based neural models that produce Potts energy tables have utility for flexible applications in protein science. CodeCode will be made publically available at https://github.com/alexjli/terminator_public

biophysics↗

bioRxiv2022

Amplified cortical neural responses as animals learn to use novel activity patterns

Cerebral cortex supports representations of the world in patterns of neural activity, used by the brain to make decisions and guide behavior. Past work has found diverse, or limited, changes in the primary sensory cortex in response to learning, suggesting the key computations might occur in downstream regions. Alternatively, sensory cortical changes may be central to learning. We studied cortical learning by using controlled inputs we insert: we trained mice to recognize entirely novel, non-sensory patterns of cortical activity in the primary visual cortex (V1) created by optogenetic stimulation. As animals learned to use these novel patterns, we found their detection abilities improved by an order of magnitude or more. The behavioral change was accompanied by large increases in V1 neural responses to fixed optogenetic input. Neural response amplification to novel optogenetic inputs had little effect on existing visual sensory responses. A recurrent cortical model shows that this amplification can be achieved by a small mean shift in recurrent network synaptic strength. Amplification would seem to be desirable to improve decision-making in a detection task, and therefore these results suggest that adult recurrent cortical plasticity plays a significant role in improving behavioral performance during learning.

neuroscience↗

bioRxiv2021

Parametric modulators of sex-biased conditioned fear responding

Pavlovian fear conditioning is a widely used behavioral paradigm for studying associative learning in rodents. Despite early recognition that subjects may engage in a variety of both conditioned and unconditioned responses, the last several decades have seen the field narrow its focus to measure freezing as the sole indicator of conditioned fear. We previously reported that female rats were more likely than males to engage in darting, an escape-like conditioned response that is associated with heightened shock reactivity. To determine how experimental parameters contribute to the frequency of darting in both males and females, we manipulated factors such as chamber size, shock intensity, and number of trials. To better capture fear-related behavioral repertoires in our animals, we developed ScaredyRat, an open-source custom Python tool that analyzes Noldus Ethovision-generated raw data files to identify Darters and quantify both conditioned and unconditioned responses. We find that like freezing, conditioned darting occurrences scale with experimental alterations. While most darting occurs in females, we find that with an extended training protocol, darting can emerge in males as well. Collectively, our data suggest that darting reflects a behavioral switch in conditioned responding that is a product of both an individual animals sex, shock reactivity, and experimental parameters, underscoring the need for careful consideration of sex as a biological variable in classic learning paradigms.

animal behavior and cognition↗

bioRxiv2020

DNA methylation landscapes of matched primary and recurrent high grade serous ovarian cancers are preserved throughout disease progression and chemoresistance.

Little is known about the role of global DNA methylation in recurrence and chemoresistance of high grade serous ovarian cancer (HGSOC). We performed whole genome bisulfite sequencing (WGBS) and whole transcriptome sequencing (RNA-seq) to establish methylation and gene expression signatures in 62 primary and recurrent tumors from 28 patients diagnosed with stage III/IV HGSOC. Eleven of these patients carried pathogenic germline BRCA1/BRCA2 mutations. Genome-wide methylation and transcriptomic features identified in primary tumors were largely preserved in matched recurrent tumors from the same patient (P-value = 7.16 x 10-7 and 1.41 x 10-3 in BRCA1/2 and non-BRCA1/2 cases respectively). Tumors from BRCA1/2 carriers displayed high levels of heterogeneity, with significantly more shared methylation changes identified between primary and recurrent tumors from non-BRCA1/2 patients, which may be related to the poorer survival we observe in HGSOCs from non-BRCA1/2 carriers (P-value = 0.0056). Partially methylated domains (PMDs) dominated the epigenetic variation across all tumors, and were more hypomethylated in BRCA1/2 than non-BRCA1/2 cases. Differential gene expression analysis identified upregulation of genes from immune pathways including antigen processing and presentation in tumors from BRCA1/2 carriers, implicating increased immune response in the improved survival observed in these patients. In summary, this study shows a previously unreported conservation of methylation and gene expression in recurrent HGSOCs. These data have implications for the possible effectiveness of epigenetic based therapies to treat both primary and recurrent ovarian cancers.

genomics↗